Support Vector Machines Project

Welcome to your Support Vector Machine Project! Just follow along with the notebook and instructions below. We will be analyzing the famous iris data set!

The Data

For this series of lectures, we will be using the famous Iris flower data set.

The Iris flower data set or Fisher's Iris data set is a multivariate data set introduced by Sir Ronald Fisher in the 1936 as an example of discriminant analysis.

The data set consists of 50 samples from each of three species of Iris (Iris setosa, Iris virginica and Iris versicolor), so 150 total samples. Four features were measured from each sample: the length and the width of the sepals and petals, in centimeters.

Here's a picture of the three different Iris types:

In [2]:
# The Iris Setosa
from IPython.display import Image
url = 'http://upload.wikimedia.org/wikipedia/commons/5/56/Kosaciec_szczecinkowaty_Iris_setosa.jpg'
Image(url,width=300, height=300)
Out[2]:
In [18]:
# The Iris Versicolor
from IPython.display import Image
url = 'http://upload.wikimedia.org/wikipedia/commons/4/41/Iris_versicolor_3.jpg'
Image(url,width=300, height=300)
Out[18]:
In [19]:
# The Iris Virginica
from IPython.display import Image
url = 'http://upload.wikimedia.org/wikipedia/commons/9/9f/Iris_virginica.jpg'
Image(url,width=300, height=300)
Out[19]:

The iris dataset contains measurements for 150 iris flowers from three different species.

The three classes in the Iris dataset:

Iris-setosa (n=50)
Iris-versicolor (n=50)
Iris-virginica (n=50)

The four features of the Iris dataset:

sepal length in cm
sepal width in cm
petal length in cm
petal width in cm

Get the data

Use seaborn to get the iris data by using: iris = sns.load_dataset('iris')

In [1]:
import pandas as pd
import numpy as np
import seaborn as sns
import matplotlib.pyplot as plt

iris = sns.load_dataset('iris')
In [2]:
iris.head()
Out[2]:
sepal_length sepal_width petal_length petal_width species
0 5.1 3.5 1.4 0.2 setosa
1 4.9 3.0 1.4 0.2 setosa
2 4.7 3.2 1.3 0.2 setosa
3 4.6 3.1 1.5 0.2 setosa
4 5.0 3.6 1.4 0.2 setosa
In [3]:
iris.info()

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 150 entries, 0 to 149
Data columns (total 5 columns):
sepal_length    150 non-null float64
sepal_width     150 non-null float64
petal_length    150 non-null float64
petal_width     150 non-null float64
species         150 non-null object
dtypes: float64(4), object(1)
memory usage: 5.9+ KB
In [4]:
iris.describe()
Out[4]:
sepal_length sepal_width petal_length petal_width
count 150.000000 150.000000 150.000000 150.000000
mean 5.843333 3.057333 3.758000 1.199333
std 0.828066 0.435866 1.765298 0.762238
min 4.300000 2.000000 1.000000 0.100000
25% 5.100000 2.800000 1.600000 0.300000
50% 5.800000 3.000000 4.350000 1.300000
75% 6.400000 3.300000 5.100000 1.800000
max 7.900000 4.400000 6.900000 2.500000
In [5]:
iris.columns
Out[5]:
Index(['sepal_length', 'sepal_width', 'petal_length', 'petal_width',
       'species'],
      dtype='object')

Let's visualize the data and get you started!

Exploratory Data Analysis

Create a pairplot of the data set. Which flower species seems to be the most separable?

In [6]:
sns.pairplot(data=iris,hue='species', palette='Dark2')
Out[6]:
<seaborn.axisgrid.PairGrid at 0x1c2b9fc7dd8>

Create a kde plot of sepal_length versus sepal width for setosa species of flower.

In [7]:
setosa = iris[iris['species'] == 'setosa']
sns.kdeplot( data=setosa['sepal_width'], data2=setosa['sepal_length'], cmap="plasma", shade=True, shade_lowest=False)

D:\Anaconda\lib\site-packages\scipy\stats\stats.py:1713: FutureWarning: Using a non-tuple sequence for multidimensional indexing is deprecated; use `arr[tuple(seq)]` instead of `arr[seq]`. In the future this will be interpreted as an array index, `arr[np.array(seq)]`, which will result either in an error or a different result.
  return np.add.reduce(sorted[indexer] * weights, axis=axis) / sumval
Out[7]:
<matplotlib.axes._subplots.AxesSubplot at 0x1c2bd732940>

Train Test Split

Split your data into a training set and a testing set.

In [8]:
from sklearn.model_selection import train_test_split
In [9]:
df_feat = iris.drop(columns=['species'], axis=1)
df_feat.head()
Out[9]:
sepal_length sepal_width petal_length petal_width
0 5.1 3.5 1.4 0.2
1 4.9 3.0 1.4 0.2
2 4.7 3.2 1.3 0.2
3 4.6 3.1 1.5 0.2
4 5.0 3.6 1.4 0.2
In [10]:
df_target = iris['species']
df_target.unique()
Out[10]:
array(['setosa', 'versicolor', 'virginica'], dtype=object)
In [27]:
#target class to numerical label
from sklearn.preprocessing import LabelEncoder

LE = LabelEncoder()
array_target = LE.fit_transform(iris['species'])

df_target = pd.DataFrame(data=array_target, columns=['target_code'])
In [29]:
df_target.head()
Out[29]:
target_code
0 0
1 0
2 0
3 0
4 0
In [19]:
X_train, X_test, y_train, y_test = train_test_split(df_feat, df_target, test_size=0.30)

Train a Model

Now its time to train a Support Vector Machine Classifier.

Call the SVC() model from sklearn and fit the model to the training data.

In [12]:
from sklearn.svm import SVC
In [13]:
model = SVC()
In [20]:
model.fit(X=X_train, y=y_train)
Out[20]:
SVC(C=1.0, cache_size=200, class_weight=None, coef0=0.0,
  decision_function_shape='ovr', degree=3, gamma='auto', kernel='rbf',
  max_iter=-1, probability=False, random_state=None, shrinking=True,
  tol=0.001, verbose=False)

Model Evaluation

Now get predictions from the model and create a confusion matrix and a classification report.

In [21]:
y_pred = model.predict(X_test)
In [16]:
from sklearn.metrics import classification_report, confusion_matrix
In [22]:
print(confusion_matrix(y_test, y_pred))

[[14  0  0]
 [ 0 12  1]
 [ 0  0 18]]
In [23]:
print(classification_report(y_pred=y_pred,y_true=y_test))

             precision    recall  f1-score   support

     setosa       1.00      1.00      1.00        14
 versicolor       1.00      0.92      0.96        13
  virginica       0.95      1.00      0.97        18

avg / total       0.98      0.98      0.98        45

Wow! You should have noticed that your model was pretty good! Let's see if we can tune the parameters to try to get even better (unlikely, and you probably would be satisfied with these results in real like because the data set is quite small, but I just want you to practice using GridSearch.

Gridsearch Practice

Import GridsearchCV from SciKit Learn.

In [25]:
from sklearn.model_selection import GridSearchCV

Create a dictionary called param_grid and fill out some parameters for C and gamma.

In [26]:
param_grid = {'C': [0.1,1, 10, 100], 'gamma': [1,0.1,0.01,0.001]}

Create a GridSearchCV object and fit it to the training data.

In [27]:
grid = GridSearchCV(SVC(),param_grid,refit=True,verbose=2)
grid.fit(X_train,y_train)

Fitting 3 folds for each of 16 candidates, totalling 48 fits
[CV] C=0.1, gamma=1 ..................................................
[CV] ................................... C=0.1, gamma=1, total=   0.0s
[CV] C=0.1, gamma=1 ..................................................
[CV] ................................... C=0.1, gamma=1, total=   0.0s
[CV] C=0.1, gamma=1 ..................................................
[CV] ................................... C=0.1, gamma=1, total=   0.0s
[CV] C=0.1, gamma=0.1 ................................................
[CV] ................................. C=0.1, gamma=0.1, total=   0.0s
[CV] C=0.1, gamma=0.1 ................................................
[CV] ................................. C=0.1, gamma=0.1, total=   0.0s
[CV] C=0.1, gamma=0.1 ................................................
[CV] ................................. C=0.1, gamma=0.1, total=   0.0s
[CV] C=0.1, gamma=0.01 ...............................................
[CV] ................................ C=0.1, gamma=0.01, total=   0.0s
[CV] C=0.1, gamma=0.01 ...............................................
[CV] ................................ C=0.1, gamma=0.01, total=   0.0s
[CV] C=0.1, gamma=0.01 ...............................................
[CV] ................................ C=0.1, gamma=0.01, total=   0.0s
[CV] C=0.1, gamma=0.001 ..............................................
[CV] ............................... C=0.1, gamma=0.001, total=   0.0s
[CV] C=0.1, gamma=0.001 ..............................................
[CV] ............................... C=0.1, gamma=0.001, total=   0.0s
[CV] C=0.1, gamma=0.001 ..............................................
[CV] ............................... C=0.1, gamma=0.001, total=   0.0s
[CV] C=1, gamma=1 ....................................................
[CV] ..................................... C=1, gamma=1, total=   0.0s
[CV] C=1, gamma=1 ....................................................
[CV] ..................................... C=1, gamma=1, total=   0.0s
[CV] C=1, gamma=1 ....................................................
[CV] ..................................... C=1, gamma=1, total=   0.0s
[CV] C=1, gamma=0.1 ..................................................
[CV] ................................... C=1, gamma=0.1, total=   0.0s
[CV] C=1, gamma=0.1 ..................................................
[CV] ................................... C=1, gamma=0.1, total=   0.0s
[CV] C=1, gamma=0.1 ..................................................
[CV] ................................... C=1, gamma=0.1, total=   0.0s
[CV] C=1, gamma=0.01 .................................................
[CV] .................................. C=1, gamma=0.01, total=   0.0s
[CV] C=1, gamma=0.01 .................................................
[CV] .................................. C=1, gamma=0.01, total=   0.0s
[CV] C=1, gamma=0.01 .................................................
[CV] .................................. C=1, gamma=0.01, total=   0.0s
[CV] C=1, gamma=0.001 ................................................
[CV] ................................. C=1, gamma=0.001, total=   0.0s
[CV] C=1, gamma=0.001 ................................................
[CV] ................................. C=1, gamma=0.001, total=   0.0s
[CV] C=1, gamma=0.001 ................................................
[CV] ................................. C=1, gamma=0.001, total=   0.0s
[CV] C=10, gamma=1 ...................................................
[CV] .................................... C=10, gamma=1, total=   0.0s
[CV] C=10, gamma=1 ...................................................
[CV] .................................... C=10, gamma=1, total=   0.0s
[CV] C=10, gamma=1 ...................................................
[CV] .................................... C=10, gamma=1, total=   0.0s
[CV] C=10, gamma=0.1 .................................................
[CV] .................................. C=10, gamma=0.1, total=   0.0s
[CV] C=10, gamma=0.1 .................................................
[CV] .................................. C=10, gamma=0.1, total=   0.0s
[CV] C=10, gamma=0.1 .................................................
[CV] .................................. C=10, gamma=0.1, total=   0.0s
[CV] C=10, gamma=0.01 ................................................
[CV] ................................. C=10, gamma=0.01, total=   0.0s
[CV] C=10, gamma=0.01 ................................................
[CV] ................................. C=10, gamma=0.01, total=   0.0s
[CV] C=10, gamma=0.01 ................................................

[Parallel(n_jobs=1)]: Done   1 out of   1 | elapsed:    0.0s remaining:    0.0s

[CV] ................................. C=10, gamma=0.01, total=   0.0s
[CV] C=10, gamma=0.001 ...............................................
[CV] ................................ C=10, gamma=0.001, total=   0.0s
[CV] C=10, gamma=0.001 ...............................................
[CV] ................................ C=10, gamma=0.001, total=   0.0s
[CV] C=10, gamma=0.001 ...............................................
[CV] ................................ C=10, gamma=0.001, total=   0.0s
[CV] C=100, gamma=1 ..................................................
[CV] ................................... C=100, gamma=1, total=   0.0s
[CV] C=100, gamma=1 ..................................................
[CV] ................................... C=100, gamma=1, total=   0.0s
[CV] C=100, gamma=1 ..................................................
[CV] ................................... C=100, gamma=1, total=   0.0s
[CV] C=100, gamma=0.1 ................................................
[CV] ................................. C=100, gamma=0.1, total=   0.0s
[CV] C=100, gamma=0.1 ................................................
[CV] ................................. C=100, gamma=0.1, total=   0.0s
[CV] C=100, gamma=0.1 ................................................
[CV] ................................. C=100, gamma=0.1, total=   0.0s
[CV] C=100, gamma=0.01 ...............................................
[CV] ................................ C=100, gamma=0.01, total=   0.0s
[CV] C=100, gamma=0.01 ...............................................
[CV] ................................ C=100, gamma=0.01, total=   0.0s
[CV] C=100, gamma=0.01 ...............................................
[CV] ................................ C=100, gamma=0.01, total=   0.0s
[CV] C=100, gamma=0.001 ..............................................
[CV] ............................... C=100, gamma=0.001, total=   0.0s
[CV] C=100, gamma=0.001 ..............................................
[CV] ............................... C=100, gamma=0.001, total=   0.0s
[CV] C=100, gamma=0.001 ..............................................
[CV] ............................... C=100, gamma=0.001, total=   0.0s

[Parallel(n_jobs=1)]: Done  48 out of  48 | elapsed:    0.2s finished
Out[27]:
GridSearchCV(cv=None, error_score='raise',
       estimator=SVC(C=1.0, cache_size=200, class_weight=None, coef0=0.0,
  decision_function_shape='ovr', degree=3, gamma='auto', kernel='rbf',
  max_iter=-1, probability=False, random_state=None, shrinking=True,
  tol=0.001, verbose=False),
       fit_params=None, iid=True, n_jobs=1,
       param_grid={'C': [0.1, 1, 10, 100], 'gamma': [1, 0.1, 0.01, 0.001]},
       pre_dispatch='2*n_jobs', refit=True, return_train_score='warn',
       scoring=None, verbose=2)

Now take that grid model and create some predictions using the test set and create classification reports and confusion matrices for them. Were you able to improve?

In [28]:
grid_y = grid.predict(X_test)
In [29]:
print(confusion_matrix(y_pred=grid_y, y_true= y_test))

[[14  0  0]
 [ 0 12  1]
 [ 0  0 18]]
In [30]:
print(classification_report(y_pred=grid_y, y_true=y_test))

             precision    recall  f1-score   support

     setosa       1.00      1.00      1.00        14
 versicolor       1.00      0.92      0.96        13
  virginica       0.95      1.00      0.97        18

avg / total       0.98      0.98      0.98        45
In [61]:
 

             precision    recall  f1-score   support

     setosa       1.00      1.00      1.00        15
 versicolor       1.00      0.93      0.96        14
  virginica       0.94      1.00      0.97        16

avg / total       0.98      0.98      0.98        45

You should have done about the same or exactly the same, this makes sense, there is basically just one point that is too noisey to grab, which makes sense, we don't want to have an overfit model that would be able to grab that.

Great Job!